Comment by unfletch | Hacker News Reader

Comment by unfletch | Hacker News Reader

As far as I can tell this might just be false. If you go to github.com (a random site which happens to include query 1.7.1 -- the latest version), open the console in either firebug or webkit inspector, and run the author's example:

    var foo = $("body");
    foo.data("mykey", "1.410");
    foo.data("mykey");

...you get your original "1.410", not 1.41 as the author reports.

Secondly, he says there are "several bug reports on the jQuery website, where people are also confused by this behavior and all of them get closed with a wontfix" but links to none of them. Linking to them should be trivial and they'd help his case (or at least make it easy for me to figure out why he sees a bug and I don't). I searched and couldn't find any.

Finally, you shouldn't use .attr() to store arbitrary data as he recommends. Hanging javascript data off of DOM nodes will cause serious memory leaks if you're not careful. That's the reason .data() exists in the first place. See the "Cicular References" section of Understanding and Solving Internet Explorer Leak Patterns: http://msdn.microsoft.com/en-us/library/Bb250448

Sorry, I shouldn't have had quotes around the 1.410

http://jsfiddle.net/kgp2H/

Think of it this way, say you have a template that is putting some data into an element like this:

<button id="fooButton" data-key="1.400">Click me to edit something</button>

Now, using .data() to select the information out, will always return 1.4 instead of what you want, which is 1.400.

You are not being serious, right?

1.410 === 1.41

Yep, latchkey, I think you might want to rethink the article. As others have pointed out, you're seeing JavaScript's behaviour, not jQuery's. And this:

  elem.data('foo', { 'a':1, 'b':2 });
  elem.data('foo'); // Returns the object

Is certainly more desirable than:

  elem.attr('foo', { 'a':1, 'b':2 });
  elem.attr('foo'); // Returns "[object Object]"

His example is unclear, but he's complaining about the behavior of conditionally casting the initial data values based on pattern matching.

   <div id="test" data-foo="1.1"> 
   $('#test').data('bar', '1.1');
   typeof $('#test').data('foo'); // number
   typeof $('#test').data('bar'); // string

If the data-foo value is user input, then converting it to a number might not be expected. E.g.

   <div data-emoticon="0.0" />

Hm, now I guess you better just delete your article. 1.410 IS 1.41, isn't it. Or am I missing something.

Then what's the problem? You stored float 1.410 and got back float 1.41. They're equal. That's not jQuery, it's javascript: http://jsfiddle.net/kgp2H/1/

Edit, after the parent was edited to include the data attribute example:

You should update your post to clarify that you're talking about data attributes. That's a significant detail and shouldn't have been left out.

Now that you've clarified, jQuery does behave as you've stated, but it's not a bug. I can see how it could be unexpected behavior to some, but it's explicitly stated in the .data() documentation:

  Every attempt is made to convert the string to a 
  JavaScript value (this includes booleans, numbers, 
  objects, arrays, and null) otherwise it is left as a 
  string. To retrieve the value's attribute as a string 
  without any attempt to convert it, use the attr() method.

http://api.jquery.com/data/#data2

That may be the documented behavior, but that doesn't make it the correct behavior. The problem is that we have a typeless language applying arbitrary type detection to strings... and it will invariably guess wrong.

A string of "1.410" is not the same as a float of 1.41. Running "1.410" full-circle through data() corrupts the input string. When you have data that has been entered by a user, you can't guarantee that it's not something that will happen to autoconvert.

This behavior is totally broken and should never be used. It will work 99.95% of the time, then fail in bizarre ways when a user types in something magic. These are the worst bugs in the world to hunt down. This really shouldn't be in jquery.

But 1.410 isn't a string, it's a float. What do you expect jQuery to do, parse the source of its caller and determine that you wrote 1.410 rather than 1.41? That's the only thing it could reasonably do, because the value it actually receives as an argument is identical in both cases.

Your example isn't very clear. http://jsfiddle.net/4bd23/1/ shows that data read from the HTML attribute may be coerced into a number, while data values set from JavaScript are always returned as the same type.

I can see why they would do this. Having to parseInt all default values wouldn't be very jQuery-like.

Reply on news.ycombinator.com